作者: Philip L. Swan
DOI:
关键词:
摘要: An image reconstruction system adaptively de-interlaces video stream content using data comparison techniques when the interlaced input does not contain pre-coded non-interlaced to conversion status data. In one embodiment, uses a signature generator which generates plurality of field values on per basis by determining region based fluctuations in pixel data, such as luminance within horizontal scan line detect changes motion. The are then analyzed determine probability that is several types has undergone conversion, pull down for film captured content. Also, independent whether occurred, analyzes streams contains paused or slow motion playback images and accordingly.